lance-format / lance-format/lance

Document and test RAM requirements

Open
#5,101 1 comment 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

ci
Dominant language
Rust
Stars
7.1k
Forks
852
Avg merge
3d 18h
Merged PRs (30d)
272

Description

For major operations we should be able to document and test what RAM is required:

  • What is the peak RAM that could be allocated during the operation
  • What items could the operation cause to be loaded into the metadata cache
  • What items could the operation cause to be loaded into the index cache

We should have tests that verify the above. In addition, these tests should verify that all RAM is returned if the caches are disabled.

Operations that need documenting:

  • Inserting data
  • Scanning data (ignoring filter / search)
  • Performing a vector search
    • For each kind of vector index
  • Utilizing a scalar index
    • For each kind of scalar index
  • Performing a merge insert
  • Deleting data
  • Running compaction
  • Training a new index
    • For each kind of index
  • Updating an index
    • For each kind of index
  • Running cleanup

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating the implementations and existing tests for inserting, scanning, searching, indexing, compaction, and cleanup. Map the peak allocation and metadata or index cache effects for each operation, then add documentation and tests that verify the measurements and confirm memory is returned when caches are disabled.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
databases, performance
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.